A rotorcraft blade health monitoring and early warning device and method based on a flexible sensor

Abstract

The application provides a rotorcraft blade health monitoring and early warning device and method based on a flexible sensor and belongs to the technical field of aircraft safety detection. Mainly comprises: the rotorcraft body, characterized in that the rotorcraft body comprises a self-sensing blade, a collection and transmission module and an upper computer; the self-sensing blade comprises a propeller blade, a flexible sensor integrated on the propeller blade and a wire for connecting the flexible sensor; the flexible sensor comprises a vibration sensor, a strain sensor, a pressure sensor, a temperature sensor and a humidity sensor; the pressure sensor is arranged on the upper and lower surfaces of the propeller blade along the length direction of the propeller blade, and the upper and lower surface arrangement sites are the same; the collection and transmission module is used for transmitting the signal collected by the flexible sensor to the upper computer; the upper computer is used for analyzing the signal and making a warning judgment, so that the safety and reliability of the rotorcraft are improved.

Description

Technical Field

[0001] This invention belongs to the field of aircraft safety detection technology, specifically relating to a rotor blade health monitoring and early warning device and method based on flexible sensors. Background Technology

[0002] With the widespread application of rotorcraft in logistics, agriculture, inspection, and urban air transportation, the health of propellers or rotors, as core power components, directly affects flight safety and performance. Traditional methods for monitoring rotor health, such as vibration analysis or motor current monitoring, suffer from indirectness and insufficient accuracy, while rigid sensors may affect the aerodynamic characteristics of the rotor and even introduce additional structural risks. Rotor health monitoring technology based on flexible sensors offers a new approach to solving these problems.

[0003] Flexible sensors, based on novel materials (such as graphene and nanofibers) and micro / nano manufacturing technologies, are characterized by their thinness, flexibility, and high sensitivity. They can conform to the complex curved surfaces of rotors and monitor multi-dimensional parameters such as strain, cracks, temperature, and vibration in real time. For example, strain sensors can detect the bending deformation of rotors, while piezoelectric sensors can capture changes in dynamic vibration frequencies, thereby indirectly identifying micro-damage within the material and promptly preventing blade cracks, fractures, or fatigue failures. This direct monitoring method significantly improves the accuracy and timeliness of fault detection while avoiding the interference of traditional sensors on aerodynamic performance.

[0004] Furthermore, the combination of flexible sensors with data acquisition and wireless transmission technologies further reduces system power consumption and weight, solves the wiring challenges of wired acquisition on high-speed rotating components, and meets the lightweight design requirements of rotorcraft. The acquired data is transmitted to airborne or ground-based computing units, and combined with machine learning algorithms, fault characteristics can be extracted from multi-source sensor data, enabling early warning and remaining life prediction of abnormal states such as rotor icing, fatigue cracks, and delamination damage. Currently, the unique advantages of this technology may become an important development direction in the field of intelligent health monitoring for rotorcraft, and in the future, it may drive the implementation of innovative applications such as "self-sensing rotors," providing key support for the safe and reliable operation of next-generation intelligent aircraft. However, how to rationally deploy flexible sensors and how to integrate data from multiple sensors for reasonable analysis to achieve multiple risk warnings are problems that urgently need to be solved. Summary of the Invention

[0005] The purpose of this invention is to provide a rotorcraft blade health monitoring and early warning device and method based on flexible sensors, which improves the ability to warn of multiple risks through the reasonable deployment of various sensors.

[0006] To achieve the above objectives, the present invention provides a rotorcraft blade health monitoring and early warning device based on flexible sensors, comprising a rotorcraft body, wherein the rotorcraft body includes self-sensing blades, a data acquisition and transmission module, and a host computer.

[0007] The self-sensing propeller blade includes a propeller blade, a flexible sensor integrated on the propeller blade, and a wire for connecting the flexible sensor; the flexible sensor includes a vibration sensor, a strain sensor, a pressure sensor, a temperature sensor, and a humidity sensor; the pressure sensor is arranged along the length direction of the propeller blade on the upper and lower surfaces of the propeller blade, and the arrangement points on the upper and lower surfaces are the same.

[0008] The acquisition and transmission module is used to transmit the signals acquired by the flexible sensor to the host computer;

[0009] The host computer is used to analyze the signal and make early warning judgments.

[0010] Furthermore, the rotorcraft body also includes a motor for driving the self-sensing propeller blades to rotate and a fairing disposed above the motor shaft, with the propeller blades fixed to the motor shaft; the data acquisition and transmission module is disposed inside the fairing, and a counterweight is provided to adjust the coaxiality of the data acquisition and transmission module and the motor shaft.

[0011] Furthermore, the vibration sensors are respectively located at the root of the propeller blade near the motor shaft and at the top of the blade away from the motor shaft; they can be integrated into the upper or lower surface of the blade at the root or top, or embedded in the middle of the blade.

[0012] The vibration sensor at the root is used to monitor the stiffness reduction signal caused by the crack; the vibration sensor at the top is used to monitor the signal change caused by torsion or flutter.

[0013] The strain sensors are arranged along the length of the propeller blade at stress concentration points (they can be integrated into the upper or lower surface of the blade or embedded in the middle of the blade) to monitor bending and tensile strain signals.

[0014] The temperature and humidity sensors are located at the airflow impact points on the leading edge of the propeller blades, and are used to jointly determine the risk of icing and high-temperature damage based on the monitoring data from the temperature, humidity, and pressure sensors.

[0015] Furthermore, the acquisition and transmission module includes an analog-to-digital conversion unit and a wireless communication unit. The analog-to-digital conversion unit is used to convert the signal acquired by the flexible sensor into an analog signal, and the wireless communication unit is used to wirelessly connect the analog-to-digital conversion unit to the host computer.

[0016] Furthermore, the host computer includes a data analysis unit and an early warning unit;

[0017] The early warning unit includes a crack risk early warning module, which issues a crack risk early warning when the vibration sensor detects that the frequency deviation of the first-order bending vibration mode exceeds a preset threshold, for example, >5%, and the strain sensor detects that the strain value exceeds the average value of the same working condition by more than 20%.

[0018] The fracture risk warning module uses a cumulative damage model to calculate fatigue life based on data from strain sensors. When the strain cycle count at a certain point reaches a preset threshold, such as 80% of the fatigue limit of the propeller blade material, and the vibration sensor detects a continuous decrease in the frequency of the corresponding area, and the pressure sensor detects a change in the pressure value at the same point exceeding a preset threshold, such as >10% / second, a fracture risk warning is issued.

[0019] The torsional risk warning module issues a torsional risk warning when the vibration sensor detects a new harmonic component and the pressure value change of two adjacent pressure sensors on the upper surface exceeds a preset threshold, for example, >15%.

[0020] The stall warning module uses a data analysis unit to calculate the aerodynamic efficiency from the pressure sensor data using a lift distribution reconstruction algorithm. When the aerodynamic efficiency drops below a preset threshold, such as >15%, and the upper surface pressure sensor detects a fluctuation frequency of 10-50Hz, and the vibration sensor detects an amplitude of 10-100Hz exceeding a preset threshold, such as >0.1g, a stall risk warning is issued.

[0021] The icing warning module will issue an icing risk warning when the temperature and humidity reach the preset icing conditions, such as temperature ≤0℃ and humidity ≥85%, and the vibration sensor detects that the amplitude of the first-order bending vibration mode is more than 30% higher than that of the healthy state.

[0022] Furthermore, the data analysis unit includes a high-temperature damage correction module, which is used to lower the warning threshold of the strain sensor when the temperature sensor detects that the surface temperature of the propeller blade is >60°C.

[0023] When the temperature is >60℃, if either or both of the following two conditions exist, the lowering range of the warning threshold of the strain sensor shall be increased. The two conditions are: the pressure sensor detects an aerodynamic curve in which the pressure distribution deviates from the normal state, and the vibration sensor detects that the frequency change rate of the first-order bending vibration mode exceeds the preset threshold.

[0024] When there is a risk of icing, the host computer controls the pressure sensor to switch to a high-frequency sampling mode of 50-150Hz to capture changes in aerodynamic parameters in the early stages of icing.

[0025] This invention also provides a method for monitoring and warning the health of rotor blades based on flexible sensors, which employs the aforementioned rotor blade health monitoring and warning device based on flexible sensors and includes the following steps:

[0026] The propeller blade signals are collected in real time using vibration sensors, strain sensors, pressure sensors, temperature sensors, and humidity sensors.

[0027] The acquired signals are transmitted to the host computer via the acquisition and transmission module.

[0028] The host computer analyzes the signals and makes early warning judgments.

[0029] Furthermore, when the vibration sensor detects a frequency shift of >5% in the first-order bending vibration mode, and the strain sensor detects a strain value exceeding 20% ​​of the average value under the same working conditions, a crack risk warning is issued.

[0030] The cumulative damage model is used to calculate fatigue life based on data from strain sensors. When the strain cycle count at a certain point reaches 80% of the fatigue limit of the propeller blade material, and the vibration sensor detects a continuous decrease in the frequency of the corresponding area, and the pressure sensor detects a pressure value change of >10% / second at the same point, a fracture risk warning is issued.

[0031] When the vibration sensor detects a new harmonic component and the pressure value change of two adjacent pressure sensors on the upper surface is greater than 15%, a torsional risk warning is issued.

[0032] The aerodynamic efficiency is calculated from the data of the pressure sensor by the lift distribution reconstruction algorithm. When the aerodynamic efficiency drops by more than 15%, and the upper surface pressure sensor detects a fluctuation frequency of 10-50Hz, and the vibration sensor detects an amplitude of more than 0.1g in the 10-100Hz frequency band, a stall risk warning is issued.

[0033] When icing conditions are detected with temperature ≤0℃ and humidity ≥85%, and the vibration sensor detects that the amplitude of the first-order bending vibration mode is increased by more than 30% compared to the healthy state, an icing risk warning is issued.

[0034] Furthermore, when the temperature sensor detects that the surface temperature of the propeller blade is greater than 60°C, the warning threshold of the strain sensor will be lowered.

[0035] When the temperature is >60℃, if either or both of the following two conditions exist, the lowering range of the warning threshold of the strain sensor shall be increased. The two conditions are: the pressure sensor detects an aerodynamic curve in which the pressure distribution deviates from the normal state, and the vibration sensor detects that the frequency change rate of the first-order bending vibration mode exceeds the preset threshold.

[0036] When there is a risk of icing, the host computer controls the pressure sensor to switch to a high-frequency sampling mode of 50-150Hz to capture changes in aerodynamic parameters in the early stages of icing.

[0037] Furthermore, the acquisition and transmission module uses a 24-bit analog-to-digital converter and a 5000× amplifier circuit to process the signal from the strain sensor; a 16-bit analog-to-digital converter is used to process the signal from the temperature sensor, and an integrated filtering circuit is used to filter out noise; a 16-bit analog-to-digital converter is used in conjunction with anti-aliasing filtering to process the signals from the vibration sensor and the pressure sensor.

[0038] In general, compared with the prior art, the above-described technical solutions conceived by this invention have the following main technical advantages:

[0039] 1. The rotor blade health detection and early warning device based on flexible sensors provided by this invention integrates a flexible sensor network directly on or inside the propeller to create a self-sensing propeller. This allows for real-time acquisition of multi-dimensional data such as strain, temperature, vibration, and pressure, achieving high-precision and high-sensitivity direct monitoring. Furthermore, through a multi-source data fusion mechanism, it can not only identify structural damage such as cracks and fractures, but also provide early warnings of environmentally induced faults such as icing and high temperatures, and even assess aerodynamic performance degradation such as airflow separation, significantly expanding the monitoring coverage and early warning capabilities.

[0040] 2. This invention constructs a multi-dimensional, cross-validated health status assessment system through the joint monitoring of multiple sensors, fundamentally solving the limitations and misjudgment risks of single-sensor monitoring. Different types of flexible sensors each perform their specific functions while complementing each other's data: strain sensors accurately capture the microscopic deformation and stress concentration of the blades, vibration sensors record dynamic frequency and amplitude changes to reflect structural integrity, temperature and humidity sensors monitor environmental induction factors (such as icing conditions) in real time, and pressure sensors assess aerodynamic performance through changes in airflow distribution. This multi-parameter collaborative acquisition allows fault identification to no longer rely on a single signal, but rather improves accuracy through "feature combinations."

[0041] 3. The multi-sensor joint monitoring of this invention significantly improves the system's robustness and anti-interference capability through data cross-verification. In complex flight environments, a single sensor may experience signal distortion due to electromagnetic interference, extreme temperatures, or mechanical vibration. However, when multiple sensors work together, the host computer algorithm can eliminate abnormal data through cross-validation. For example, if a strain sensor experiences a signal jump due to localized wear, but the data from its adjacent strain sensors remains stable, and the vibration and pressure sensors do not detect corresponding anomalies, the algorithm can determine that the signal is invalid interference, avoiding false alarms. Conversely, when multiple sensors simultaneously capture correlated features (such as a sudden increase in strain + vibration frequency shift + abnormal pressure distribution), the reliability of fault judgment can be enhanced, ensuring the reliability of the warning. This "multi-signal verification" mechanism allows the system to maintain stable detection accuracy even under complex operating conditions, providing higher safety redundancy for blade health monitoring.

[0042] 4. This invention adopts a modular design, which is highly compatible and has low maintenance costs. This solution does not require changes to the existing propeller structure design and can be directly integrated into the propeller blades during the manufacturing process. It is suitable for various rotorcraft platforms, and in subsequent maintenance, only damaged blades and data acquisition and transmission modules need to be replaced, which greatly reduces the maintenance cost of the monitoring system. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the overall structure of a rotor blade health detection and early warning device based on flexible sensors.

[0044] Figure 2 This is a schematic diagram of the installation of a self-sensing propeller based on a flexible sensor.

[0045] Figure 3 This is a schematic diagram of the data acquisition and transmission module.

[0046] Figure 4 This is a structural breakdown diagram of the self-sensing blade.

[0047] Figure 5 This is a schematic diagram of the installation of the data acquisition and transmission module.

[0048] Figure 6 This is a schematic diagram of a method for detecting and warning the health of rotor blades based on flexible sensors.

[0049] In all the accompanying drawings, the same reference numerals are used to denote the same elements or structures, wherein:

[0050] 1-Fairing, 2-Data acquisition and transmission module, 201-Wireless transmission unit, 202-Sensor network interface, 203-Analog-to-digital conversion unit, 204-Counterweight, 3-Shock-absorbing pad, 4-Self-sensing propeller blade, 401-Flexible sensor, 402-Wire, 403-Propeller blade, 5-Motor, 6-Host computer. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0052] Please see Figure 1-6 This invention provides a rotorcraft propeller health detection and early warning device and method based on flexible sensors. Through a closed-loop logic of "real-time sensing of the self-sensing propeller → precise conversion by the data acquisition and transmission module → intelligent analysis and early warning by the host computer," it achieves full-process monitoring of the propeller's health status. It mainly comprises three parts: a data acquisition and transmission module 2, a self-sensing propeller 4 integrating a flexible sensor network, and an onboard computing unit (host computer 6). This technology has advantages such as high precision and real-time performance, enabling high-precision self-sensing capability for rotorcraft propeller blades.

[0053] Specifically, the rotor blade health monitoring and early warning device based on flexible sensors includes a rotor blade body, which includes a self-sensing blade 4, a data acquisition and transmission module 2, and a host computer 6.

[0054] The self-sensing propeller blade 4 includes a propeller blade 403, a flexible sensor 401 integrated on the propeller blade 403, and a wire 402 for connecting the flexible sensor; the flexible sensor 401 includes a vibration sensor, a strain sensor, a pressure sensor, a temperature sensor, and a humidity sensor; the pressure sensor is arranged along the length direction of the propeller blade 403 on the upper and lower surfaces of the propeller blade 403, and the arrangement points on the upper and lower surfaces are the same;

[0055] The acquisition and transmission module 2 is used to transmit the signal acquired by the flexible sensor 401 to the host computer 6;

[0056] The host computer 6 is used to analyze the signal and make early warning judgments.

[0057] Specifically, the type and number of flexible sensors 401 in the self-sensing blade 4 are determined according to the blade monitoring requirements, covering various types such as strain, temperature, humidity, pressure, and vibration. In practical applications, it is necessary to analyze the possible damage types and common operating conditions of the rotor blade 403 to determine the specific types of sensors required. If it is necessary to monitor damage such as blade deformation, cracks, and fractures, these damages will change the stress-strain state and vibration characteristics of the rotor blade 403, so vibration and strain sensors are selected. If it is necessary to monitor conditions such as icing and high-temperature damage in extreme environments, the blade needs to sense high / low temperature and humidity environmental conditions as well as airflow pressure distribution, so temperature, humidity, and pressure sensors are selected. If it is necessary to monitor conditions such as strong winds and flow separation, these environments will cause abnormal aerodynamic loads on the blade, which will lead to changes in pressure distribution and increased vibration, so vibration and pressure sensors are selected.

[0058] To reduce weight and minimize the impact on blade performance, the sensor is primarily a thin film fabricated using micro-nano manufacturing technology. Its pins are connected to fine wires via conductive silver paste or low-temperature solder, and the output signal is in the form of resistance, voltage, capacitance, etc. The sensor and the wires together form a flexible sensing network.

[0059] Specifically, to ensure that the dynamic balance of the rotor is not affected during the rotation of motor 5, the sensors installed on the blades of the same motor should be symmetrically distributed about the motor shaft. The distribution of sensors on the blades should be designed and adjusted according to the size, shape, and material of the propeller blades 403. Before designing the sensor distribution, professional software is used to simulate the vibration modes, fluid dynamics, and thermodynamics of the propeller blades 403 to obtain information such as stress concentration points, large deformation areas, vibration modes, and areas prone to icing.

[0060] Furthermore, vibration and strain sensors are integrated to address structural faults such as fatigue cracks, torsional chatter, and fracture deformation. Vibration sensors are integrated at the root (closer to motor 5) and tip (farthest from motor 5) of propeller blade 403. The root vibration sensors are positioned at the antinodes of the first-order bending vibration mode, monitoring frequency shifts within a 100Hz range before and after the natural frequency band to capture stiffness reduction caused by cracks. The tip vibration sensors are positioned at sensitive points of higher-order vibration modes, focusing on amplitude abrupt changes in higher-order frequency bands (specific frequency bands determined by simulation data) to identify early characteristics of torsional or chattering. Simultaneously, strain sensors are distributed along the blade's length, embedded at internal stress concentration points (such as root transition fillets and abrupt changes in the middle section, determined by simulation results). A full-bridge circuit is used to improve measurement accuracy and eliminate interference from environmental factors such as temperature, enabling real-time monitoring of bending and tensile strain. When the strain value at a certain point exceeds the historical average under the same working conditions by 20%, and the strain gradient of adjacent sensors simultaneously exceeds the historical average under the same working conditions by 20%, it can be determined that the local deformation has intensified. At this time, the vibration sensor data is combined to judge the blade vibration condition and monitor whether there are signs of fracture in blade 403.

[0061] To address environmentally induced faults such as icing and high-temperature damage, temperature and humidity sensors are integrated at the airflow impact point on the upper surface of the blade leading edge (the primary area of ​​water droplet impact and air friction, derived from simulation data). Simultaneously, pressure sensors are intermittently integrated into the aerodynamically sensitive area on the lower surface of the blade to reconstruct the rotor flow field. The temperature sensor accuracy must be ±0.5℃, the humidity sensor resolution ≥1%RH, and the pressure sensor resolution higher than ±10Pa. Icing risk is assessed when "temperature ≤0℃ + humidity ≥85%" and atypical pressure fluctuations occur. In high-temperature environments, the reliability of high-temperature damage assessment is enhanced by combining the difference between the blade surface temperature and the ambient temperature with deviations in pressure distribution from the normal aerodynamic curve. Furthermore, the captured airflow load data provides aerodynamic load correction for the strain sensor's warning threshold, improving the accuracy and robustness of environmentally induced fault monitoring.

[0062] Furthermore, a thin-film pressure sensor is deployed on the upper surface of blade 403, at the same location as the pressure sensor on the lower surface, covering the core area where lift is generated. The lift distribution is reconstructed by changes in pressure gradient; when local lift suddenly drops, the host computer alerts to a decrease in aerodynamic efficiency. The frequency of pressure fluctuations caused by airflow separation (irregular fluctuations of 10-50Hz are flow separation signals) is monitored to provide early warnings of strong winds and flow separation. Data from both types of pressure sensors corroborate each other, avoiding the limitations of monitoring only a single area.

[0063] Specifically, the flexible sensors are all thin-film sensors fabricated using micro-nano manufacturing technology and integrated with the blades using resin adhesive: surface sensors are completely bonded to the curved surface of the blade; embedded sensors are tightly bonded to fiber-reinforced materials, curing together with the fiber material during the layer-by-layer bonding process of the blade fibers to ensure strain transfer efficiency ≥90%. During network construction, to avoid blade damage due to high temperatures, conductive silver paste or low-temperature solder is used to connect to ultra-fine wires 402 with a diameter of 0.1mm. The wires are routed along pre-reserved channels inside the blades, avoiding sensitive areas such as vibration antinodes and stress concentration points identified in simulation results. The exit points are fixed with sealant to prevent detachment during high-speed rotation and to avoid disrupting the aerodynamic shape. If necessary, the surface of the flexible sensor 401 needs to be covered with a polyimide encapsulation layer to withstand airflow erosion and UV aging.

[0064] Specifically, after integrating the flexible sensor 401 and the wire 402 into the propeller blade 403 to form the self-sensing blade 4, the self-sensing blade 4 is wired to the data acquisition and transmission module 2. The data acquisition and transmission module 2 is fixed above the shaft of the motor 5, which facilitates the data acquisition circuit to receive the data transmitted by the sensor, while not affecting the dynamic balance of the motor 5. The data acquisition and transmission module 2 is divided into an analog-to-digital conversion unit 203, a sensor network interface 202, and a wireless transmission unit 201. The analog-to-digital conversion unit 203 performs customized signal processing for different types of sensors: the mV-level signal output by the strain sensor needs to be processed by a 24-bit ADC (sampling rate 1kHz) and a 5000× amplifier circuit; the resistance signal of the temperature sensor is converted into a linear voltage signal (10mV / ℃) by a high-precision constant current source, which can be processed by a 16-bit ADC (sampling rate 10Hz) and integrated with a low-pass filter circuit to filter out high-frequency noise; the resistance signal of the pressure sensor is converted into a standard voltage signal by a dedicated signal conditioning chip, and then processed together with the voltage signal directly output by the vibration sensor by a 16-bit ADC (sampling rate 2kHz) with anti-aliasing filtering to ensure that the dynamic signal is not distorted; the capacitance signal output by the humidity sensor is processed into an acquireable voltage signal by a capacitance-to-voltage conversion circuit and then acquired by a 16-bit ADC (sampling rate 10Hz).

[0065] The wireless transmission unit 201 dynamically switches modes based on data volume. Vibration and pressure sensors, with higher data throughput (approximately 200kB / s), utilize WiFi (approximately 200kB / s) for transmission; temperature, humidity, and strain sensors, with lower data throughput (approximately 10kB / s), utilize Bluetooth (power consumption <50mW) for transmission. The two modes are automatically switched via commands from the host computer 6, balancing transmission efficiency and energy consumption. During installation, the acquisition and transmission module 2 requires adjustment of its center of gravity using counterweights 204 to ensure a coaxiality error ≤0.1mm with the motor shaft, preventing additional vibration during rotation. The acquisition and transmission module 2 is integrated within the rectifier 1 above the motor, using silicone pads for shock absorption (resonance frequency <20Hz) to ensure stable operation at propeller speeds of 1000-8000rpm. The power supply system prioritizes wireless power (transmission efficiency ≥70%). If lithium battery power is used, the battery level must be sufficient for at least 8 hours of continuous operation, and the host computer monitors the battery level in real-time, triggering a low-battery warning when it falls below 20%.

[0066] Specifically, after receiving the data from the acquisition and transmission module 2, the host computer 6 (airborne computing unit) uses artificial intelligence algorithms such as machine learning and neural networks to analyze the sensor data and issue timely warnings.

[0067] For vibration data, the Fast Fourier Transform (FFT) is used to extract frequency features. Combined with the neural network model, the frequency spectrum under healthy conditions is compared. When the first-order frequency deviation at the root is greater than 5%, and the strain sensor at the corresponding position synchronously detects a strain value that exceeds the average value under the same working conditions by more than 20%, it is determined that crack initiation has occurred, and a crack risk warning is issued.

[0068] If the vibration sensor at the top detects new harmonic components at higher frequencies, and the pressure sensor detects asymmetrical pressure distribution on the upper surface (i.e., the pressure value change of two adjacent pressure sensors is greater than 15%), a torsion risk warning will be issued.

[0069] For pressure data, aerodynamic efficiency is calculated using a lift distribution reconstruction algorithm. When the aerodynamic efficiency drops by more than 15%, and the upper surface pressure sensor detects an irregular fluctuation frequency (flow separation signal) of 10-50Hz, while the vibration sensor detects an abnormal amplitude (>0.1g) in the 10-100Hz frequency band, a stall risk warning is triggered.

[0070] For temperature and humidity data, when icing conditions of "temperature ≤ 0℃ + humidity ≥ 85%" are detected, if the amplitude of the vibration sensor in the low-frequency band (10-50Hz) increases by more than 30% compared to the healthy state, icing is confirmed and a de-icing command is triggered. At the same time, if the pressure sensor simultaneously detects an asymmetrical lift distribution between the left and right blades (e.g., pressure difference > 20%), the risk of icing is further confirmed, and the early warning can be strengthened.

[0071] In high-temperature environments (blade surface temperature > 60℃), the warning threshold of the strain sensor is lowered to varying degrees based on real-time data from the temperature sensor (the degree of lowering is dynamically adjusted according to temperature and material type, with a preferred reduction range of 5%-20%). Simultaneously, the fracture risk assessment due to material strength decay is dynamically corrected by referencing the frequency change rate of the vibration sensor (> 0.5 Hz / min). For strain data, a cumulative damage model is used to calculate fatigue life. When the strain cycle count at a certain point reaches 80% of the material's fatigue limit, and the vibration sensor detects a continuous decrease in frequency in the corresponding area (> 2 Hz / hour), and the pressure sensor simultaneously shows a local pressure gradient anomaly (pressure value change > 10% / second), a high fracture risk is identified, and the flight control system is immediately notified to execute an emergency landing procedure.

[0072] The multi-sensor joint monitoring of this invention can cover more complex fault scenarios, realizing an upgrade from "single damage detection" to "full life cycle health management". For progressive damage such as fatigue cracks, the frequency shift detected by the vibration sensor needs to be synchronously correlated with the strain cycle accumulation data of the strain sensor and the local pressure change of the pressure sensor in order to accurately locate the crack and assess the propagation speed. For aerodynamic stall risk, the pressure sensors on the upper and lower surfaces reconstruct the lift distribution through gradient changes, combined with the 10-50Hz irregular fluctuations captured by the vibration sensor, to provide a more comprehensive basis for early warning.

[0073] Preferably, if the fracture point is located exactly in the middle of the sensor, the sensor will be physically damaged along with the broken blade, causing the sensor's signal to be interrupted. In this case, a multi-dimensional mechanism can be designed to avoid sensor failure. For example, sensor redundancy design can be implemented, using a distributed layout (such as embedding multiple points along the blade length). Even if one sensor is damaged due to breakage, sensors in other locations can still detect anomalies (such as sudden strain changes in adjacent areas or abrupt changes in vibration frequency), enabling cross-verification. Abnormal sensor signals are used as early warning signals. The host computer 6 can use the "sudden interruption of a single sensor signal" itself as a fault characteristic, combined with data from other sensors (such as strain peak values ​​and vibration mode distortion from adjacent sensors), to comprehensively determine if a breakage has occurred, avoiding missed detections due to a single sensor failure. The host computer 6 dynamically adjusts the sampling rate based on the sensor status. When the algorithm identifies "abnormal vibration signal but normal strain signal," it automatically instructs the acquisition module to increase the sampling rate of the strain sensor (from 1kHz to 2kHz) and simultaneously activates the backup strain sensor (redundant design) to verify the authenticity of the vibration signal. When "high risk of icing" is detected, it automatically instructs the pressure sensor to switch to "high-frequency sampling mode" (from 10Hz to 100Hz) to capture subtle changes in aerodynamic parameters during the initial stages of icing.

[0074] In summary, the present invention proposes a rotor blade health detection and early warning device and method based on flexible sensing technology. By integrating flexible sensors into the self-sensing blade, combined with the acquisition module and the onboard computing unit (host computer), it can realize real-time monitoring of the blade health status, and predict possible damage such as breakage, deformation, and icing of the blade through the intelligent algorithm of the host computer, thereby improving the safety and reliability of the rotorcraft.

[0075] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A rotorcraft blade health monitoring and early warning device based on flexible sensors, comprising a rotorcraft body, characterized in that, The rotorcraft body includes self-sensing blades, a data acquisition and transmission module, and a host computer. The self-sensing propeller blade includes a propeller blade, a flexible sensor integrated on the propeller blade, and a wire for connecting the flexible sensor; the flexible sensor includes a vibration sensor, a strain sensor, a pressure sensor, a temperature sensor, and a humidity sensor; the pressure sensor is arranged along the length direction of the propeller blade on the upper and lower surfaces of the propeller blade, and the arrangement points on the upper and lower surfaces are the same. The acquisition and transmission module is used to transmit the signals acquired by the flexible sensor to the host computer; The host computer is used to analyze the signal and make early warning judgments; The rotorcraft body also includes a motor for driving the self-sensing propeller blades to rotate and a fairing disposed above the motor shaft, wherein the propeller blades are fixed to the motor shaft; the data acquisition and transmission module is disposed inside the fairing, and a counterweight is provided for adjusting the coaxiality of the data acquisition and transmission module and the motor shaft; The vibration sensors are respectively installed at the root of the propeller blades on the side closer to the motor shaft and at the top of the blades on the side farther from the motor shaft. The vibration sensor at the root is used to monitor the stiffness reduction signal caused by the crack; the vibration sensor at the top is used to monitor the signal change caused by torsion or flutter. The strain sensors are arranged along the length of the propeller blade at stress concentration points to monitor bending strain and tensile strain signals. The temperature and humidity sensors are located at the airflow impact points on the leading edge of the propeller blades, and are used to jointly determine the risk of icing and high-temperature damage based on the monitoring data from the temperature, humidity, and pressure sensors. The host computer includes a data analysis unit and an early warning unit; The early warning unit includes a crack risk early warning module, which issues a crack risk early warning when the vibration sensor detects that the frequency deviation of the first-order bending vibration mode exceeds a preset threshold and the strain sensor detects that the strain value exceeds the average value under the same working conditions by more than 20%. The fracture risk warning module uses a cumulative damage model to calculate fatigue life based on data from strain sensors. When the strain cycle count at a certain point reaches a preset threshold, the vibration sensor detects a continuous decrease in the frequency of the corresponding area, and the pressure sensor detects a change in pressure value at the same point exceeding a preset threshold, a fracture risk warning is issued. The torsional risk warning module issues a torsional risk warning when the vibration sensor detects new harmonic components and the pressure value change of two adjacent pressure sensors on the upper surface exceeds a preset threshold. The stall warning module uses a data analysis unit to calculate the aerodynamic efficiency from the pressure sensor data through a lift distribution reconstruction algorithm. When the aerodynamic efficiency drops below a preset threshold and the upper surface pressure sensor detects a fluctuation frequency of 10-50Hz, and the vibration sensor detects a low-frequency amplitude exceeding a preset threshold, a stall risk warning is issued. The icing warning module will issue an icing risk warning when the temperature and humidity reach the preset icing conditions and the vibration sensor detects that the amplitude of the first-order bending vibration mode is more than 30% greater than that of the healthy state. The data analysis unit includes a high-temperature damage correction module, which is used to lower the warning threshold of the strain sensor when the temperature sensor detects that the surface temperature of the propeller blade is >60°C. Furthermore, when the temperature is >60℃, if either or both of the following two conditions exist, the lowering range of the warning threshold of the strain sensor will be increased. The two conditions are: the pressure sensor detects an aerodynamic curve in which the pressure distribution deviates from the normal state, and the vibration sensor detects that the frequency change rate of the first-order bending vibration mode exceeds the preset threshold. When there is a risk of icing, the host computer controls the pressure sensor to switch to a high-frequency sampling mode of 50-150Hz to capture changes in aerodynamic parameters in the early stages of icing.

2. The rotor blade health monitoring and early warning device based on flexible sensors according to claim 1, characterized in that, The acquisition and transmission module includes an analog-to-digital conversion unit and a wireless communication unit. The analog-to-digital conversion unit is used to convert the signals acquired by the flexible sensor into analog signals, and the wireless communication unit is used to wirelessly connect the analog-to-digital conversion unit to the host computer.

3. A method for monitoring and early warning the health of rotor blades of a rotary-wing aircraft based on flexible sensors, characterized in that, The rotor blade health monitoring and early warning device based on flexible sensors as described in any one of claims 1-2 includes the following steps: The propeller blade signals are collected in real time using vibration sensors, strain sensors, pressure sensors, temperature sensors, and humidity sensors. The acquired signals are transmitted to the host computer via the acquisition and transmission module. The host computer analyzes the signals and makes early warning judgments.

4. The method for monitoring and early warning of rotor blade health based on flexible sensors according to claim 3, characterized in that, When the vibration sensor detects that the frequency deviation of the first-order bending vibration mode exceeds a preset threshold, and the strain sensor detects that the strain value exceeds 20% of the average value under the same working conditions, a crack risk warning is issued. The cumulative damage model is used to calculate fatigue life based on data from strain sensors. When the strain cycle count at a certain point reaches a preset threshold, the vibration sensor detects a continuous decrease in the frequency of the corresponding area, and the pressure sensor detects a change in pressure value at the same point exceeding a preset threshold, a fracture risk warning is issued. When the vibration sensor detects a new harmonic component and the pressure value change of two adjacent pressure sensors on the upper surface exceeds a preset threshold, a torsional risk warning is issued. The aerodynamic efficiency is calculated from the data of the pressure sensor by the lift distribution reconstruction algorithm. When the aerodynamic efficiency drops below a preset threshold and the upper surface pressure sensor detects a fluctuation frequency of 10-50Hz, and the vibration sensor detects a low-frequency amplitude exceeding a preset threshold, a stall risk warning is issued. When the temperature and humidity reach the preset icing conditions, and the vibration sensor detects that the amplitude of the first-order bending vibration mode is more than 30% greater than that in the healthy state, an icing risk warning will be issued.

5. The method for monitoring and early warning of rotor blade health based on flexible sensors according to claim 4, characterized in that, When the temperature sensor detects that the surface temperature of the propeller blade is greater than 60°C, the warning threshold of the strain sensor will be lowered. When the temperature is >60℃, if either or both of the following two conditions exist, the lowering range of the warning threshold of the strain sensor shall be increased. The two conditions are: the pressure sensor detects an aerodynamic curve in which the pressure distribution deviates from the normal state, and the vibration sensor detects that the frequency change rate of the first-order bending vibration mode exceeds the preset threshold. When there is a risk of icing, the host computer controls the pressure sensor to switch to high-frequency sampling mode to capture changes in aerodynamic parameters in the early stages of icing.

6. The method for monitoring and early warning of rotor blade health based on flexible sensors according to claim 3, characterized in that, The data acquisition and transmission module uses an analog-to-digital converter and an amplifier circuit to process the signals from the strain sensor; it also uses an analog-to-digital converter to process the signals from the temperature sensor and integrates a filter circuit to remove noise; and it uses an analog-to-digital converter in conjunction with an anti-aliasing filter to process the signals from the vibration sensor and the pressure sensor.

Images